Although sufferers of panic disorder often present to primary care clinics, the condition often goes undetected in this setting. Delays in its recognition carry a heavy burden of functional impairment and cost. At our Anxiety Disorders Clinic, we derived a 4-item self-report screen, called the Panic Disorder Screen (PDS), from the 16-item Anxiety Sensitivity Index (ASI) which distinguishes panic disorder patients from other anxiety disorder patients: high PDS or ASI scores indicate panic disorder. This application proposes a pilot study to investigate the diagnostic accuracy and utility of the PDS in a primary care setting using individuals presenting to a general medicine clinic. The specific aims: 1) to evaluate the criterion-related validity of the PDS and the ASI, and 2) to evaluate the sensitivity, specificity, and positive and negative predictive value of the PDS and the ASI for panic disorder in subjects presenting to a medical clinic, compared with diagnosis established by structured clinical interview. Patients registering at the Cornell Medical Practice, a medical clinic serving a diverse urban population, will complete the ASI when they complete a general medical health form. Individuals who score high on the ASI or its PDS subscale and a random selection of those who score low are recruited for further evaluation. Subjects participate in a structured interview to determine the presence or absence of psychiatric disorders. Factor structure, consistency, and test-retest reliability of the PDS and the ASI will be evaluated. Receiver Operating Characteristic curves will be generated to select a cutscore which optimizes sensitivity and specificity. The predictive value of ASI and PDS will help evaluate its utility as a screen for panic disorder. This exploratory study will potentially facilitate more open and fruitful collaboration between psychiatrists and internists. For internists, it promises the ability to more actively consider the diagnosis of panic disorder. For psychiatrists, it opens the exploration of the relationship among subject variables, somatic complaint variables, PDS scores, and structured clinical interview findings.